April 2010
BACKGROUNDER
Abundant Shale Gas
Resources: Some
Implications for Energy
Policy
St ephen P. A. Brow n, St even A. G abriel, and Ruud Eggi ng
1616 P St. NW Washington, DC 20036 202-328-5000
www.rff.org
A
b
s
tr
a
c
t
Abundant Shale Gas Resources:
Some Implications for Energy
Policy
Stephen P.A. Brown, Steven A. Gabriel, and Ruud
Egging
According to recent assessments, the United States has considerably more
recoverable natural gas in shale formations than was previously thought. Such a
development raises the possibility of a shift in U.S. energy consumption toward
natural gas. To examine how the apparent abundance of natural gas might affect
U.S. energy markets and the role of natural gas in climate policy, we model five
scenarios— reflecting different perspectives on natural gas availability, the
availability of competing resources, and climate policy—through 2030. We find that
more abundant natural gas supplies result in greater natural gas use in most sectors
of the economy. We further find that natural gas could serve as a bridge fuel to a
low-carbon future, but only if appropriate low-carbon policies are in place.
Key Words: natural gas markets, shale gas resources, climate
policy, energy policy
This background paper is one in a series developed as part of the
Resources for the Future and National Energy Policy Institute project
entitled “Toward a New National Energy Policy: Assessing the Options.”
This project was made possible through the support of the George Kaiser
Family Foundation.
© 2010 Resources for the Future. All rights reserved. No portion of this paper may
be reproduced without permission of the authors.
Background papers are research materials circulated by their authors for purposes of
information and discussion. They have not necessarily undergone formal peer review.
Contents 1. Introduction
......................................................................................................................... 1
2. U.S. Natural Gas Markets
.................................................................................................. 3
2.1 Natural Gas Consumption
............................................................................................. 3
2.2 Natural Gas Supply
....................................................................................................... 4
2.3 Natural Gas Resources in Shale Formations
................................................................. 5
2.4 Natural Gas
Infrastructure............................................................................................. 5
2.5 Alaskan Natural Gas and Pipeline
................................................................................ 6
3. How NEMS-RFF Represents U.S. Natural Gas Markets
............................................... 6
3.1 Natural Gas Supply and Demand
.................................................................................. 7
3.2 Natural Gas Transmission and Distribution
.................................................................. 9
4. Assessing the Implications of More Abundant Natural Gas
........................................... 9
4.1 The Baseline Case (Scenario 1)
.................................................................................. 10
4.2 Implications of Abundant Natural Gas Supply (Scenario 2)
...................................... 11
4.3 How Natural Gas Supply Affects Carbon Policy
....................................................... 12
4.3.1 Low-Carbon Policy without Abundant Natural Gas (Scenario 3)
......................12
4.3.2 Low-Carbon Policy with Abundant Natural Gas (Scenario 4)
...........................13
4.3.3 How Abundant Natural Gas AffectsLow-Carbon Policy
...................................13
4.4 Limits on Nuclear and Renewable Power Generation (Scenario 5)
........................... 14
4.5 The Incidence of Low-Carbon Policy
......................................................................... 15
4.6 The Potential Effects of Intervening Mandates
.......................................................... 17
5. Some Potential Issues with NEMS-RFF
......................................................................... 18
5.1 Fuel Substitution and Natural Gas and Crude Oil Prices
............................................ 18
5.2 Potential for Bottlenecks in Natural Gas Pipelines
..................................................... 20
5.3 Alaska Natural Gas Production
................................................................................... 20
5.4 Adequacy of U.S. LNG Import Terminals
.................................................................. 21
5.5 The Adequacy of NEMS-RFF
.................................................................................... 21
6. Uncertainties about U.S. Natural Gas Markets
............................................................. 22
6.1 Shale Gas Uncertainty and the Transition to a Low-Carbon Future
........................... 22
6.2 Elimination of Oil and Gas Tax Preferences
.............................................................. 22
6.3 Moratoria Lands
.......................................................................................................... 23
6.4 Coalbed Methane
........................................................................................................ 23
6.5 Gas Hydrates
............................................................................................................... 24
6.6 Canadian Exports
........................................................................................................ 24
6.7 Mexican Imports
......................................................................................................... 25
7. Summary and
Conclusions............................................................................................... 25
Tables and Figures
................................................................................................................ 27
Appendix A. The Five
Scenarios.......................................................................................... 38
Appendix B. A Comparison of Natural Gas Market
Models........................................... 39
Rice World Gas Trade Model
........................................................................................... 40
World Gas Model
.............................................................................................................. 40
GASMOD
......................................................................................................................... 41
GASTALE
........................................................................................................................ 41
GRIDNET
......................................................................................................................... 42
ICF’s GMM
...................................................................................................................... 43
Appendix B References
.................................................................................................... 43
References
.............................................................................................................................. 45
Resources for the Future Brown, Gabriel, and Egging
1. Introduction Recent assessments suggest that the United States has
considerably more recoverable
natural gas in shale formations than was previously thought, as new drilling technologies
dramatically lowered recovery costs. The apparent abundance of natural gas raises the
possibility of a substantial shift in U.S. energy consumption toward natural gas. At the
same time, many are looking to natural gas as a bridge fuel to a low-carbon future
because its use yields carbon dioxide (CO) emissions that are about 45 percent lower
per British thermal unit (Btu) than coal and 30 percent lower than oil. Such a transition
seems particularly attractive in the electric power sector if natural gas were to displace
coal.
The possibility of more abundant natural gas raises a number of questions that we
seek to answer in the present exercise. How might more abundant natural gas affect the
fuel mix? Will it create a market-driven reduction in CO2 emissions? Does it lower the cost
of policies to reduce CO emissions? Some ancillary questions are: What are the
implications for natural gas use when nuclear and renewable power generation are
limited? How is natural gas use affected if a renewable portfolio standard is used in
conjunction with a cap-and-trade system?
1To
assess how the apparent abundance of natural gas might affect U.S. energy
markets and the role of natural gas in climate policy, we compare five scenarios that
reflect different perspectives on natural gas availability, the availability of competing
resources, and climate policy. We modeled these scenarios, which run through 2030,
using NEMS-RFF. The scenarios
Stephen Brown is a nonresident fellow at Resources for the Future, Steven Gabriel is an associate
professor of civil engineering at the University of Maryland, and Ruud Egging is a Ph.D. student at the
University of Maryland. The authors thank Peter Balash, Tina Bowers, Kara Callahan, Joel Darmstadter,
Bob Fri, Less Goudarzi, Mary Haddican, Kristin Hayes, Hill Huntington, Tony Knowles, Alan Krupnick, Jan
Mares, Karen Palmer, Ian Parry, Michael Schaal, Phil Sharp, Sharon Showalter, Dana Van-Wagener,
Margaret Walls, John Weyant, and Frances Wood for helpful discussions and comments.
1 The National Energy Modeling System (NEMS) is a computer-based, energy-economy market
equilibrium
modeling system for the United States developed by the U.S. Department of Energy. NEMS-RFF is a
version of NEMS developed by Resources for the Future (RFF) in cooperation with OnLocation, Inc.
NEMS-RFF projects market-clearing prices and quantities across a number of energy markets, subject
to assumptions about macroeconomic and financial developments, world energy market conditions,
demographics, resource availability
Abundant Shale Gas Resources:
Some Implications for Energy
Policy
Stephen P.A. Brown, Steven A. Gabriel, and Ruud
Egging
1
2
2
*
*
Resources for the Future Brown, Gabriel, and Egging
reflect different perspectives on natural gas resources in shale formations, the adoption of
lowcarbon policies, and the availability of nuclear and renewable power generation.
By comparing these scenarios, we assess how the relative abundance of natural gas
might affect its consumption and its potential to reduce CO2 emissions. We find that more
abundant natural gas supplies result in greater natural gas use in most sectors of the
economy. More importantly, we find that with appropriate carbon policies in place—such as
a cap-and-trade system or a carbon tax—natural gas can play a role as a bridge fuel to a
low-carbon future. The role of natural gas as a transition fuel to a low-carbon future could be
enhanced if the use of nuclear and renewable power proves to be limited, or it could be
reduced if renewable portfolio standards are used to supplement cap-and-trade policies.
Nonetheless, we find that having low-carbon policies in place is essential if natural
gas is to serve as a bridge to a low-carbon future. Without such policies, more abundant
natural gas does not reduce CO2 emissions. Although greater natural gas resources reduce
the price of natural gas and displace the use of coal and oil, they also boost overall energy
consumption and reduce the use of nuclear and renewable energy sources for electric
power generation. As a result, projected CO2 emissions are almost 1 percent higher. With a
carbon cap-and-trade system in place, however, we find that greater natural gas
supplies can help meet carbon-reduction goals. With more abundant natural gas, the use of
natural gas in electricity generation increases significantly and overall natural gas
consumption remains robust, which lessens slightly the burden on other measures to
reduce CO emissions. In addition, the price of CO22 allowances falls slightly, which lessens
the economic cost of reducing CO2 emissions somewhat. It is this ability to lower the costs
of climate policy that makes natural gas an attractive bridge fuel to a low-carbon future.
From a broader perspective, however, our analysis suggests that the most
cost-effective means for reducing CO2 emissions depends greatly on projected resource
availability and technology changes, both of which are highly uncertain. If policymakers are
to develop meaningful and cost-effective policies for controlling CO2 emissions, they must
develop policies that are robust across different projected futures.
and costs, the cost and performance characteristics of energy technologies, and behavioral and technological
choice criteria.
2
Resources for the Future Brown, Gabriel, and Egging
2. U.S. Natural Gas Markets Natural gas is used widely throughout the U.S.
economy—in industry, residences and
commercial establishments, and to generate electric power. About 90 percent of U.S.
natural gas consumption is met by domestic production. The remainder is imported from
Canada or from other countries as liquefied natural gas (LNG). Increased production from
shale formations is expected to reduce reliance on LNG imports in future years. Throughout
the United States, a substantial system of pipelines connects natural gas producers with
their consumers.
2.1 Natural Gas Consumption Natural gas plays an important role in U.S. energy use. It
accounted for nearly 25 percent
2of total U.S. energy consumption in 2008. Only petroleum product consumption accounted
for more (Figure 1). In contrast with oil and coal (the use of which are concentrated in
transportation and electric power, respectively), natural gas is used across a variety of
sectors in the U.S. economy (Figure 2). The industrial sector is the largest user, accounting
for nearly 35 percent of total natural gas consumption.3 The electric power sector accounts
for nearly 30 percent. The residential sector is the second-largest end-use sector,
accounting for more than 20 percent of total natural gas consumption. Natural gas is also
used in the commercial sector, and a small amount of natural gas is used in the
transportation sector. The use of natural gas to generate electric power increased sharply in
the late 1990s and
early 2000s. The growth was driven by changes in regulation, new technology, relatively
abundant natural gas, lower capital costs relative to those for coal and nuclear power
plants, and environmental advantages over coal and oil. As the electric power sector was
restructured, some regulatory advantages for natural gas use in power generation were
eliminated. Those developments led to idle generation capacity and
weaker-than-anticipated growth in natural gas use in the electric power sector.
Owing to its use for space heating and in peaking plants for electric power generation,
the demand for natural gas is highly seasonal and weather related. The seasonal peak
occurs during the winter months, but spikes occur when there is a strong demand for winter
heating or summer air conditioning, with the latter supplied by natural gas–fired electricity.
The seasonal and
This figure includes lease and plant fuel.
This figure includes natural gas used as a fuel for pipelines used in its own
transportation.
3
2
3
Resources for the Future Brown, Gabriel, and Egging
weather-related swings in consumption are met by natural gas in storage. Natural gas
production and imports are relatively steady throughout the year. In a typical year, natural
gas is put into storage over the summer months from late spring to late fall and is withdrawn
over the winter months from late fall to early spring.
2.2 Natural Gas Supply As shown in Figure 3, U.S. natural gas supplies come from a
variety of sources, including
4domestic production, imports through pipelines from Canada, and LNG from a number of
countries.5 Currently, almost 99 percent of U.S. gas consumption comes from domestic
sources and Canada. Under baseline projections from NEMS-RFF, which were made with
conservative assumptions about shale gas resources, that percentage is expected to fall to
slightly less than 97 percent in 2030 as the United States imports more LNG. Over the
same time horizon, U.S. natural gas production is expected to continue a
transition to unconventional sources, such as tight sands, shale, and coalbed methane.
Arctic production sites in Alaska and Canada’s Mackenzie Delta are also thought to be
important future sources of North American natural gas. In the more distant future, coalbed
methane and gas hydrates may contribute significant shares of U.S. natural gas production.
A recent study by the National Petroleum Council (NPC; 2007) and the Annual
Energy Outlook 2009 produced by the Energy Information Agency (EIA) both underscore
the likely transition in North American natural gas production over the next 20-plus years.
According to NPC, the production of natural gas from conventional resources in the lower
48 states will decline while production from unconventional resources in the lower 48,
resources in the Arctic (such as Alaska and Canada’s Mackenzie Delta), and/or imports of
LNG will increase. EIA expects that unconventional sources of natural gas will account for
more than half of U.S. production by 2030.
4 LNG is produced by cooling normally gaseous
methane to about –260°F. In this liquid form, methane takes up 1/600th of the space and can be shipped
from one location in the world to another. On the receiving side, it can be shipped to its destination in special
trucks or regasified and added to the pipeline grid.
5 The United States also exports natural gas to Mexico, and that is projected to continue under all of the
scenarios examined.
4
Resources for the Future Brown, Gabriel, and Egging
2.3 Natural Gas Resources in Shale Formations The impact of shale gas resources
on the U.S. market could be more substantial than
these projections show. In its Annual Energy Outlook 2009, EIA placed U.S. shale
resources at 269.3 trillion cubic feet with total U.S. natural gas resources of 1,759.5 trillion
cubic feet. In contrast, Navigant Consulting (2008) finds that U.S. shale gas resources
could be as high as 842 trillion cubic feet, and the Potential Gas Committee (PGC; 2009)
provides an estimate of 615.9 trillion cubic feet. As shown in Figure 4, these shale gas
resources are widely distributed throughout the United States.
Despite the substantially higher estimates of shale gas resources, the likely supply
profiles remain highly uncertain because the industry has relatively little experience in
producing natural gas from such formations. In addition, the U.S. Environmental Protection
Agency (EPA) is taking steps toward regulating the hydraulic fluids that have been important
to enhancing the production of shale gas and boosting the estimates of recoverable
resources from shale gas formations (Obey 2009). Industry sources variously say that EPA
regulation could have no effect, could slightly increase the cost of shale gas production, or
could completely shut it down.
2.4 Natural Gas Infrastructure As shown in Figure 5, a substantial transmission and
distribution system connects natural
gas producers with their customers. Natural gas produced at the wellhead, known as wet
gas, contains a mixture of hydrocarbons and impurities. Wet gas is typically consolidated
in a collection system. At that point, natural gas processing plants remove impurities and
other hydrocarbons so that nearly pure methane (CH4) is obtained.6 The processed gas,
known as dry gas, is moved from the producing fields to market in
natural gas pipelines. Most of the larger customers, such as electric utilities and industry,
are served directly by pipelines. Residential and commercial customers who use relatively
small quantities of natural gas are served by local distribution companies, which are
themselves served by the pipelines.
LNG, another source of natural gas for the United States, is imported in special
ships through eight terminals on the East Coast and along the Gulf of Mexico (Table 1).
At these
6 The hydrocarbons removed include propane and
butane, which are typically sold as part of the petroleum product chain.
5
Resources for the Future Brown, Gabriel, and Egging
terminals, regasification facilities return the LNG to a gaseous state and inject it into the
pipeline system. Together, the existing terminals offer a capacity of 4.2 trillion cubic feet
annually, nearly 20 percent of current U.S. consumption.
2.5 Alaskan Natural Gas and Pipeline Alaska is a potential source of natural gas in North
America. PGC (2009) estimates that
Alaska has 193 trillion cubic feet of natural gas resources, accounting for about 18 percent of
conventional, non-shale natural gas resources in the United States. Obviously, a pipeline
needs to be built from the potential producing regions in Alaska to the lower 48 if these
natural gas resources are to reach a sizable U.S. market.
A number of competing proposals are being considered. The TransCanada and
ExxonMobil consortium are proposing a pipeline (Figure 6) that would extend some 1,700
miles from Alaska to connections in Alberta. To reach markets in the lower 48, the Alaska
natural gas would need to travel as much as an additional 1,500 miles using existing or new
pipelines. The BP and ConocoPhillips consortium is considering a route through Yukon and
British Columbia along the Alaska Highway and then to Alberta. Other proposed routes
include moving the natural gas across Alaska to an LNG export terminal in Valdez and
through the Mackenzie River Valley in the Northwest Territory. The latter route is attractive
because it could also serve natural gas production in the Mackenzie Delta, another
potentially large source of Arctic natural gas (Figure 7).
3. How NEMS-RFF Represents U.S. Natural Gas Markets To evaluate the effects of
more abundant natural gas from shale deposits, we use NEMSRFF. The National Energy Modeling System (NEMS) is a computer-based,
energy-economy market equilibrium modeling system for the United States developed by
the U.S. Department of Energy. NEMS-RFF is a version of NEMS developed by Resources
for the Future (RFF) in cooperation with OnLocation, Inc. NEMS-RFF projects
market-clearing prices and quantities across a number of energy markets, subject to
assumptions about macroeconomic and financial developments, world energy market
conditions, demographics, resource availability and costs, cost and performance
characteristics of energy technologies, and behavioral and technological choice criteria.
NEMS-RFF consists of a set of models (or modules) that represent energy supply
and demand for all fuels and all demand sectors in the U.S. economy. In NEMS-RFF, all
major fuel
6
Resources for the Future Brown, Gabriel, and Egging
supply markets, conversion sectors, and end-use sectors of the energy system are
represented, as are macroeconomic activity and links to international markets. The system
represents energy end-use markets for the residential, commercial, industrial, and
transportation sectors with 12 regions in the United States, 2 in Canada, and 1 in Mexico.8
Electric power generation is represented as a conversion sector using natural gas and other
energy sources to provide electricity to the four end-use sectors. 7
3.1 Natural Gas Supply and Demand The natural gas market is represented as a
market-clearing relationship established
between the modules representing natural gas demand and those representing natural gas
supply. The demand side includes modules for the residential, commercial, industrial, and
transportation sectors and electric power generation. The supply side includes an oil and
natural gas supply module, a petroleum market module, and an international energy
module. Natural gas consumption and production are linked through a natural gas
transmission and distribution module. The demand modules are driven in part by a
macroeconomic activity module.
Together, the modules use 11 key drivers to establish market-clearing prices and
quantities for U.S. natural gas markets in every year through 2030. The demand side
includes macroeconomic activity, such as industrial output; population growth; prices for
competing fuels and technologies; technological progress, learning curves, and cost
improvements; capital investment and retirement rates for existing capital stock; and
consumer choice preferences and hurdle rates of return. The supply side includes ultimate
recoverable resources and finding rates; natural gas supply and demand in Canada; supply
curves of LNG; and trigger prices for new major pipelines and LNG regasification terminals.
The model reflects normal energy market dynamics in which the short-run
responses to changes in price are much weaker than the long-term responses. On the
supply side, short-run production decisions are largely dominated by existing reserves,
though production-to-reserve
7 The United States is divided into extended census
divisions that include the normal census regions in addition to splits between the South Atlantic region and
Florida; the Mountain region and Arizona/New Mexico; and the Pacific region with California, Alaska, and
Hawaii each handled separately.
8 The electricity sector is treated as an end-use sector for natural gas
demand.
7
Resources for the Future Brown, Gabriel, and Egging
9ratios
will rise with prices. Over the long run, producers have the time to develop
resources into reserves and then production, which accounts for the stronger long-run
response. NEMS-RFF has a detailed bottom-up representation of oil and natural gas
production in
North America. Exploration, development, and production costs in individual producing
regions are the major determinants of the U.S. domestic supply of natural gas. Trends in
these costs are projected with historical data. Higher prices stimulate production by
increasing production from existing fields and through increased investment in more costly
areas. Changes in tax policy, incentives, and resource accessibility also affect natural gas
investment and production.
On the demand side, the differences between short-run and long-run responses are
largely achieved by linking energy choices to the existing capital stock. In the short run,
natural gas consumption is dominated by such factors as the existing stock of equipment,
heating and cooling degree days, and economic activity. As existing capital stock is retired
and new investment reshapes the capital stock, end users are more able to change the
composition and quantity of their energy consumption.
The electric power sector is modeled purely on the basis of cost minimization.
Consequently, the sector shows considerable flexibility in fuel choices, and it responds
readily to changes in fuel prices. In the residential, commercial, industrial, and transportation
sectors, however, NEMS-RFF explicitly models the technology choice process. In these
sectors, the use of natural gas is very unresponsive to the differential between natural gas
prices and those for other fuels, even in the long run.
The lack of response in the residential and commercial sectors results from those
modules severely limiting consumer choices. For replacement equipment, the choice of fuel
is primarily determined by the fuel that was used in the equipment being retired. For
residential housing, the same-fuel share is 95 percent. This restriction may reflect the view
that choice is curtailed by the lack of the necessary natural gas distribution systems in
regions that are not currently served very well by natural gas.
The lack of response in the industrial sector is the result of the corresponding module
showing that nearly every facility that can switch to natural gas from another fuel has done
so the
9 Resource exhaustion imposes no limits on
U.S. natural gas production over the time horizon of the model. At current U.S. natural gas production
levels, the initial gas resource estimates plus the additional shale resources estimated by the PGC would
last another 100 years.
8
Resources for the Future Brown, Gabriel, and Egging
lack of response in the transportation sector is the result of that module lacking any
economic determinants in substituting compressed natural gas (CNG) or LNG for diesel
fuel in long-haul trucking or CNG for gasoline in auto transportation.
3.2 Natural Gas Transmission and Distribution To complete the North American
natural gas market in NEMS-RFF, the Natural Gas
Transmission and Distribution Module simulates the natural gas flows between the 15
regions in the United States, Canada, and Mexico, meeting the demand of the five
consumption sectors in all U.S. regions while taking into account restrictions in the
transmission and distribution networks. Supplies, flows, and delivered prices are determined
in a low-demand and a highdemand period. Pipeline tariffs have two parts: a fixed
reservation charge and an operational usage fee.
New interregional pipeline capacity is built as the costs can be recovered in later years. 10
Under the model’s tariff structure, little excess capacity is built and pipeline capacity
constraints can affect consumption during periods of high natural gas demand.
4. Assessing the Implications of More Abundant Natural Gas To assess how more
abundant natural gas might affect energy markets, CO2 emissions, and the role of natural
gas as a bridge fuel to a low-carbon future, we compare five scenarios developed with
NEMS-RFF that reflect different perspectives on natural gas availability, climate policy, and
the availability of competing resources. Two of these scenarios are business-as-usual
(BAU) cases that assume that the United States adopts no new policies to reduce CO 11 2
emissions. The first scenario uses what now seem to be conservative estimates of shale
gas resources of 269.3 trillion cubic feet that date from 2007, but were used recently in
EIA’s Annual Energy Outlook 2009. The second uses newer, more optimistic estimates of
615.9 trillion cubic feet developed by PGC (2009).
10 Investments along major new pipeline routes,
such as those from Alaska and the Canadian Mackenzie Delta, occur after a sufficiently high natural gas price
is reached in the lower 48 states to justify the investment. As a result of restrictions placed on the model, the
Alaskan pipeline can be built at earliest in 2020, the Mackenzie pipeline in 2014 at earliest. Expansions of
these pipelines can occur after an additional increase in gas prices.
11 EIA has released its preliminary Annual Energy Outlook 2010, which places shale resources between the
estimates used for Annual Energy Outlook 2009 and those of PGC. In contrast, Navigant Consulting (2008)
offers a mean estimate of 274 trillion cubic feet with a maximum of 842 trillion cubic feet.
9
Resources for the Future Brown, Gabriel, and Egging
The third and fourth scenarios are policy scenarios and are used to examine how
different assumptions about natural gas resources affect the implementation of a low-carbon
policy. Both assume that the United States adopts a cap-and-trade policy with CO2
emissions targets similar to those in the American Clean Energy and Security Act (H.R.
2454), proposed by Representatives Henry Waxman and Edward Markey, and to those
proposed by the Obama administration prior to the U.N. climate conference in Copenhagen.
Like the first scenario, Scenario 3 uses conservative estimates of U.S. shale resources of
natural gas. Scenario 4 uses the more optimistic estimates. A comparison of these scenarios
with the first two allows us to assess how important enhanced natural gas supplies may be
to the development of climate policy.
The fifth scenario examines how the use of natural gas fares under a low-carbon
policy when recourse to nuclear and renewable power is limited or more costly. It is built on
Scenario 4 but further assumes that the use of nuclear power is limited and that renewable
energy sources have higher costs. The comparison of this scenario with the second and
fourth allows us to assess the extent to which abundant natural gas or conservation might be
used meet climate policy objectives when recourse to nuclear and renewable power
generation is limited.
4.1 The Baseline Case (Scenario 1) The baseline scenario (Scenario 1) represents
business as usual.12 It is based on EIA’s Annual Energy Outlook 2009 as revised in April
2009 to include energy provisions in the stimulus package. It also advances the
implementation of more stringent corporate average fuel economy standards from 2020 to
2016, as required by an Obama executive order issued in May 2009. Following estimates
from Annual Energy Outlook 2009, the baseline scenario assumes U.S. shale natural gas
resources of 269.3 trillion cubic feet with total U.S. natural gas resources of 1,759.5 trillion
cubic feet.
U.S. natural gas consumption is projected to grow insignificantly from 2008 to 2030
under Scenario 1. Moreover, the projection shows natural gas consumption falling through
2014, particularly in electricity generation because recent changes in regulatory actions are
expected to result in renewable energy, coal, and nuclear power continuing to crowd out
natural gas. After 2014, the adjustment to regulatory change is mostly complete, and the
electric power sector shows increasing use of natural gas.
12 This scenario is identified as Core 1 in Krupnick
et al. (2010), which is considered the master report for the RFF study in which the present exercise is
included.
10
Resources for the Future Brown, Gabriel, and Egging
4.2 Implications of Abundant Natural Gas Supply (Scenario 2) EIA’s estimates of U.S.
natural gas supplies in Annual Energy Outlook 2009 date from
2007. Given recent technological changes in extracting natural gas from shale deposits
through hydraulic fracturing, the cost of recovering natural gas from shale formations has
dropped significantly. In addition, more natural gas is recoverable. Reflecting these
developments, PGC estimates U.S. shale gas resources at 615.9 trillion cubic feet. PGC
estimates are carefully researched and well documented, and previous PGC estimates
have been used in EIA analyses.
13The
second scenario is identical to the first, except that it substitutes the PGC estimates
for U.S. shale gas resources. It assumes that more natural gas can be produced from
each well. This approach yields both lower production costs and increased shale gas
resources, with total U.S. natural gas resources boosted to 2,106.1 trillion cubic feet.
Together, Scenarios 1 and 2 offer a wide range of plausible estimates for shale gas
14resources and reflect the considerable uncertainty about the supply or production profiles
from shale gas formations. With this change, U.S. natural gas production is much stronger
than in Scenario 1 (Figure 8). U.S. natural gas production shows only a mild decline from
2009 to 2014, with steady gains coming after that. With additional natural gas supplies,
natural gas prices are lower—with the sharpest differences projected after 2025 (Figure 9).
By 2030, the projections show Henry Hub natural gas prices more than 20 percent lower
than in Scenario 1 (falling from $8.81 per million Btu to $6.86). 15 Lower natural gas prices
delay the development of a natural gas pipeline from Alaska to the lower 48 states by three
years and lead to 1.0 trillion cubic feet less offshore production in 2030. Even with such
secondary effects, the responsiveness of U.S. natural gas production to changes in price is
considerably greater. With more abundant supply, natural gas consumption is sharply
higher than in Scenario
1. The biggest difference in natural gas use between Scenarios 1 and 2 is found in the
electric power sector, where natural gas consumption is 22.5 percent higher (Table 2). Most
of the gains
13 The more conservative approach of
increasing basin size or areas would capture increased resources but not reduced costs.
14 Near New Orleans, Henry Hub handles the highest volume of natural gas of any U.S. transportation node
and is close to the largest concentration of natural gas producing regions in the country. Henry Hub is a
commonly used pricing point for natural gas in the United States, and its price movements are highly
correlated with the NYMEX futures price, which is the most widely traded natural gas contract in the world.
15 The model’s implied supply elasticity for 2030 increases from 0.27 to
0.93.
11
Resources for the Future Brown, Gabriel, and Egging
in natural gas use in the electric power sector come from substitution for other energy
sources (Figure 10). Some of the gains come from slightly increased electricity use
brought about by lower electricity prices.
Perhaps surprisingly, enhanced natural gas supplies yield nearly 1 percent higher CO
emissions in 2030 compared with Scenario 1. Lower natural gas prices leads natural gas to
displace coal, which reduces CO22 emissions, but lower prices also encourage the
displacement of some zero-carbon (nuclear and renewable) electric power sources. In
addition, complex market interactions reduce projected prices for other energy resources
and boost total energy consumption (by a little more than 1 percent in 2030). Together, these
energy market changes increase CO2 emissions. These findings suggest that greater shale
gas resources by themselves do not lead the market toward reduced CO 2 emissions.
4.3 How Natural Gas Supply Affects Carbon Policy The third and fourth
scenarios represent a cap-and-trade policy without and with
enhanced natural gas resources, respectively. The CO216 emissions targets are similar to
those in H.R. 2454, but with only one billion metric tons in offsets instead of two billion.
4.3.1 Low-Carbon Policy without Abundant Natural Gas (Scenario 3)
17Scenario
3 shows the effects of adopting a low-carbon policy without abundant natural gas
resources. As shown in Figure 11, this scenario shows reduced U.S. natural gas
consumption compared to Scenario 1. Coal and oil use are also reduced, but substantial
increases are found in nuclear and renewable power generation. Overall energy
consumption is 5.4 percent lower in 2030, with electric power generation 8.2 percent lower.
Energy consumption is lower in all sectors of the economy, and CO2 emissions in 2030 fall
from 6.2 billion tons in Scenario 1 to 4.8 billion tons in Scenario 3.
In 2030, U.S. natural gas consumption is 1.7 trillion cubic feet (7.1 percent) lower
than in Scenario 1. For the same year, domestic natural gas production is 1.6 trillion cubic
feet (7.0 percent) less. Imports are reduced slightly. As shown in Table 3, natural gas
consumption is
16 Offsets are emissions reductions achieved
outside the sectors covered by the cap-and-trade program, either domestically or internationally.
Proposed legislation would allow regulated entities to purchase offsets as an alternative to reducing
emissions.
17 Scenario 3 is the same as the central cap-and-trade policy (Core 2) discussed in Krupnick et al.
(2010).
12
Resources for the Future Brown, Gabriel, and Egging
lower in all major sectors of the economy. The biggest reduction in natural gas use is in the
electric power sector, where both natural gas and coal are displaced by conservation and
gains in nuclear and renewable power generation (Figure 12).
4.3.2 Low-Carbon Policy with Abundant Natural Gas (Scenario 4) With abundant
natural gas supplies (Scenarios 2 and 4), the implementation of a cap-andtrade program yields a greater overall reduction in energy use than would occur without
abundant supplies (Scenarios 1 and 3). Because greater natural gas supplies foster higher
overall energy consumption and higher CO2 emissions in the absence of policy intervention,
bigger reductions in energy use are required to meet the CO2 emissions targets (about 4.8
billion tons in 2030). Nonetheless, a comparison of Scenario 4 with Scenario 2 shows that
natural gas production and consumption are only slightly reduced by the introduction of a
low-carbon policy when natural gas supplies are abundant (Figure 13).
In the process, the use of natural gas for electric power generation rises, although its
use falls in other sectors (Table 4). The opportunities for fuel switching between natural gas
and other fuels are not quite as plentiful in residential, commercial, and industrial sectors
(which together account for about 65 percent of natural gas use), so emissions reductions in
these sectors depend more heavily on energy conservation. Consequently, implementation
of a low-carbon policy reduces natural gas consumption in these sectors.
With abundant natural gas supplies, the share of natural gas in electric power
generation in 2030 rises from 23.9 to 27.8 percent when the low-carbon policy is
implemented (Figure 14). In contrast, without additional shale gas resources, the same
low-carbon policy yields a decline in the share of natural gas in electric power generation,
from 19.4 percent to 18.8 percent (Table 5). With more abundant, less expensive natural
gas supplies, power generation from nuclear and renewable sources loses some
advantages, and the use of coal falls by more.
In short, plentiful natural gas supplies mean that policies to reduce CO2 emissions will yield
a market-driven substitution of natural gas for other fuels in the electric power sector. In
other sectors, which show limited substitution possibilities between natural gas and other
energy sources, plentiful natural gas supplies do not have much effect on the means to
reduce CO2 emissions.
4.3.3 How Abundant Natural Gas AffectsLow-Carbon Policy Beyond its greater use in
electric power generation, the importance of abundant natural
gas to a low-carbon policy is found through a comparison of the prices for the carbon
allowances
13
Resources for the Future Brown, Gabriel, and Egging
under the various scenarios. The price of the carbon allowance rises from $18.61 per metric
ton of carbon in 2012 to $67.26 in 2030 under Scenario 3.1918 Under Scenario 4, the price of
the carbon allowance is slightly lower, rising from $18.49 per metric ton of carbon in 2012 to
$66.83 in 2030. The lower allowance prices translate into an avoidance of costs to the
economy that amounts to about $30 million in 2012 and rises to about $300 million in 2030.
Overall, welfare gains are modest over the period for which policy is modeled
(2012–2030)—about $1 billion in present discounted value terms. In short, abundant natural
gas creates a bridge to a low-carbon future under a cap-andtrade system that is not supplemented with a renewable portfolio standard or other
government mandates. As Table 6 shows, the combination of abundant natural gas and a
cap-and-trade system yields more natural gas consumption than is found under the
baseline scenario that reflects neither abundant natural gas nor a cap-and-trade system.
Economic theory suggests that other carbon-pricing systems, such as carbon taxes, would
yield substantially similar results.
4.4 Limits on Nuclear and Renewable Power Generation (Scenario 5) As reported by
Weyant (2009), the Energy Modeling Forum 22 (EMF 22), a bi-annual
conference representing the work of 18 energy-modeling teams, is in the process of
examining the effects of climate change mitigation policies on energy use. The study
generally finds that natural gas is an important fuel under U.S. policies to sharply reduce CO
emissions by 2050. Several factors drive the increased use of natural gas in the EMF 22
projections for the United States. Through the increased use of electric and plug-in hybrid
vehicles, electric power displaces petroleum in the transportation sector, which increases
overall electricity consumption and the potential for natural gas use in electric power
generation. In addition, the use of nuclear and renewable power generation is constrained in
the EMF 22 analysis, which increases reliance on natural gas in the power sector to meet
overall CO22 emissions standards. With abundant natural gas, we find that a policy to reduce
CO2 emissions yields results similar to those of the EMF 22 study for natural gas
consumption but through somewhat different avenues. Clearly, part of the difference is the
time horizon used for analysis. In our fourth scenario, plug-in hybrids penetrate only slightly
into new automobile sales by 2030, which
These prices and all others are in 2007 dollars.
19 Discounted to 2010 using an interest rate of 5
percent.
14
18
Resources for the Future Brown, Gabriel, and Egging
contributes to a very slight shift toward electricity use in the transportation sector—but not
by nearly as much the EMF 22 study shows for 2050. In addition, none of our four scenarios
limits the use of renewable power generation, and the 50-gigawatt limit we placed on
additions to nuclear power capacity might be seen as too optimistic for 2030, even though it
is nonbinding in Scenario 4.
20Scenario
5 is designed to more thoroughly examine how limiting the deployment of nuclear
and renewable power generation might affect the role of natural gas in climate change
policy. The scenario is constructed by modifying Scenario 4. A limit of 30 gigawatts is placed
on additions to nuclear power and the growth rate of renewable power generation is
restricted. These methods of restricting capacity did not affect the cost of each technology.
21As
shown in Figure 15, limits on nuclear and renewable power yield higher natural gas use
than is projected under either Scenario 2 or 4. Scenario 5 amplifies much of what is found in
Scenario 4. Increased natural gas use in the electric power sector boosts the price of natural
gas and yields reductions in natural gas use in other sectors (Table 7). In contrast to
Scenario 4, however, the net effect is increased natural gas consumption over Scenario 2.
As expected, nuclear and renewable energy play smaller roles in the electric power sector
(Figure 16). Perhaps surprisingly, the use of coal to produce electricity is also reduced.
When zero-emissions technologies are less available, it becomes more difficult to meet
emissions targets. Cheap and dirty coal is disadvantaged by the reduced availability of
zero-emissions technologies, and its use must be reduced. The advantage shifts to
moderately clean but more expensive fuels, such as natural gas.
4.5 The Incidence of Low-Carbon Policy The question of who bears the costs often
drives policy. In our particular case, the
questions focus on which side of the natural gas market bears the costs of low-carbon
policies,
20 We also assume that the use of carbon
capture and sequestration (CCS) is limited to the two gigawatts found under Scenario 4. This assumption
is made to prevent NEMS-RFF from shifting toward CCS when we assume tighter restrictions on nuclear
power and increased costs for renewable power generation.
21 The limits on nuclear and renewable power and CCS used in Scenario 5 also increase somewhat the
estimated costs of compliance with the cap-and-trade policy over that found with Scenario 4. Under Scenario
5, the price of the carbon allowance rises from $20.28 per metric ton of carbon in 2012 to $73.32 in 2030. For
Scenario 4, the comparable prices are $18.49 and $66.83 for 2012 and 2030, respectively. These differences
reflect the higher costs of reducing CO2 emissions when technology is less available.
15
Resources for the Future Brown, Gabriel, and Egging
and the extent to which more abundant natural gas resources alter the incidence of a
low-carbon policy.
Comparisons of the projected price paths with and without carbon policies reveal
who bears the costs of the policies. The extent to which producers and natural gas users
downstream bear the costs can be determined by comparing the natural gas prices they
face under the carbon policy to those they face without the carbon policy in place. With the
carbon policy in place, producers see a reduction in the Henry Hub price of natural gas.
Downstream natural gas users pay the Henry Hub price plus the costs of the necessary
CO2 emissions allowances. As shown by a comparison of Scenarios 1 and 3 in Figure 17,
the cost of a carbon policy
is shared almost equally between the natural gas producers and those using natural gas
downstream from Henry Hub when natural gas resources in shale formations are not
assumed to be abundant. Under Scenario 1, the price of natural gas at Henry Hub is $8.81
per million Btu in 2030. Under Scenario 3, the price of natural gas at Henry Hub is $7.99
per million Btu at Henry Hub in 2030 and $10.05 per million Btu when the cost of the CO 222
emissions allowances is added. When shale gas resources are assumed to be abundant,
however, the cost of the carbon
policy is shifted away from producers and toward those using natural gas downstream from
Henry Hub, as is shown by a comparison of Scenarios 2 and 4. Under Scenario 2, the price
of natural gas at Henry Hub is $6.86 per million Btu at Henry Hub in 2030. Under Scenario
4, the price of natural gas at Henry Hub is $6.67 per million Btu at Henry Hub in 2030 and
$8.71 per million Btu when the cost of the CO223 emissions allowances is added. The
difference in the incidence of the carbon policy owes to the greater implied elasticity of
natural gas supply that is found when shale gas resources are assumed to be more
abundant. Evaluating the incidence of carbon policy by comparing the differences between
Scenarios 1 and 3 on one hand and Scenarios 2 and 4 on the other may not provide a
complete picture. Those using natural gas downstream from Henry Hub do absorb a greater
share of the costs of a low-carbon policy when natural gas is more abundant, but the price
paid for natural gas
22 We use EIA’s average
CO2 emissions factors for natural gas to apply the allowance.
23 In general, the incidence of policies affecting natural gas markets depends on the elasticities of natural gas
demand and supply. The greater the elasticity of demand, the less the cost of the policy will be borne by
those who use natural gas. The greater the elasticity of supply, the less the cost of the policy will be borne by
natural gas producers.
16
Resources for the Future Brown, Gabriel, and Egging
plus the CO2 emissions allowances is lower than when natural gas is less abundant.
Abundant natural gas simply makes the price of natural gas lower, and that dominates
the results.
4.6 The Potential Effects of Intervening Mandates None of the scenarios we
examine consider the use of mandates that favor a specific
technology, but Böhringer and Rosendahl (2009) and Palmer and Sweeney (2010) show
that the addition of a renewable portfolio standard to a cap-and-trade system can
substantially alter market outcomes and increase costs. Both efforts find that the addition of
mandates for such zero-emissions energy sources to a cap-and-trade system can
disadvantage natural gas in favor of a combination of the cleanest and dirtiest technologies.
The Palmer and Sweeney work can be compared directly to ours because their study
is part of the same group of studies in which the present exercise is included. It uses
NEMS-RFF to examine the addition of a renewable portfolio standard in the electric power
sector to a cap-andtrade system for CO2 emissions. One of their scenarios adds a
renewable portfolio standard in the electric power sector to what we identify here as
Scenario 3, a case with a cap-and-trade policy and low estimates of the natural gas resource
base.
Palmer and Sweeney find that the addition of a renewable portfolio standard in the
electric power sector to an economy-wide cap-and-trade program reduces U.S. natural
gas consumption. As might be expected, natural gas consumption is reduced only in
electric power generation. A slight reduction in natural gas prices pushes up natural gas
consumption in the other major sectors. The net effect is a reduction in natural gas
consumption.
Palmer and Sweeney also find that the addition of a renewable portfolio standard to
a cap-and-trade program has very little effect on the use of coal to generate electricity. In
contrast, Böhringer and Rosendahl find that the use of a renewable portfolio standard in
conjunction with a cap-and-trade program increases coal use relative to that under a
cap-and-trade program alone. These differences may owe to differing responses to prices
between the models and/or the scale at which the programs are implemented.
Because the supply of natural gas is considerably more elastic under the scenarios
with more abundant shale gas resources, the addition of a renewable portfolio standard to a
cap-andtrade program is likely to show more dramatic effects on natural gas use under an
abundant shale gas scenario. Such a response also could lead to an increase in the use of
coal for electric power generation in comparison to that found with a cap-and-trade program
alone.
17
Resources for the Future Brown, Gabriel, and Egging
Although Böhringer and Rosendahl and Palmer and Sweeney reach slightly different
conclusions about the use of coal, their findings point in similar directions. Driven by
mandates rather than pricing, zero-emissions technologies no longer face market
competition from moderately clean fuels. To the extent that goals for CO 2 emissions are
achieved through the mandated use of clean and costly technologies, the use of cheap and
dirty technologies need not be reduced by as much. The end result is a combination of
mandates and market incentives that disadvantage natural gas and yield higher overall
costs for a given reduction in CO emissions. If mandates are unavoidable, it may be best to
design them broadly to include a wide range of options for reducing CO2 emissions. 2
5. Some Potential Issues with NEMS-RFF The use of NEMS-RFF to assess the effects of
more abundant shale gas resources raises a
number of potential issues about the robustness of the model’s projections. Prominent
among such issues is the pricing of natural gas and petroleum products and the potential
for fuel switching between these fuels. Additional possible issues include pipeline capacity
for the transportation of natural gas, the development of a pipeline from Alaska to the
lower 48 states, potential LNG exports from Alaska, and the adequacy of U.S. LNG import
terminals.
5.1 Fuel Substitution and Natural Gas and Crude Oil Prices A small body of research
with recent contributions by Villar and Joutz (2006), Brown
and Yücel (2008b), and Hartley et al. (2008) shows that natural gas prices have moved
with oil prices over the long run, probably as a result of interfuel substitution between
natural gas and petroleum products. Short-run deviations in natural gas prices from the
long-run relationship with crude oil prices is the result of seasonal variation, weather,
production disruptions, deviations of natural gas inventories from the normal seasonal
pattern, and other factors. As these intervening factors return to normal, however, the
price of natural gas follows an errorcorrection process in which the price of natural gas
gradually adjusts toward its long-run relationship with that for crude oil.
As Figure 18 shows, several of the different methods suggested by Brown and Yücel
(2008b) for relating natural gas prices to those for crude oil show that the two prices have
moved together historically. These methods include Brown and Yücel’s econometric
estimates of the long-run relationship between natural gas and crude oil prices, and two
rules of thumb that Brown and Yücel report are used in the oil patch. Under these rules of
thumb, the expected price of natural gas is obtained by dividing the price of crude oil by
either 6 or 10.
18
Resources for the Future Brown, Gabriel, and Egging
In contrast with the historical relationship, the baseline scenario developed with
NEMSRFF shows projected natural gas prices that are well below that suggested by the
crude oil prices projected in the baseline. In addition, the baseline case shows U.S. natural
gas consumption moderating through 2014.
At issue with these projections are the potential for competition between natural gas
and refined products and the projected growth of the U.S. industrial sector. An examination
of NEMS-RFF shows that most of the potential for natural gas substitution for other fuels in
the model occurs in the electric power sector where very little oil is used. In contrast, most
oil is used in the transportation sector where very little natural gas is used. Only the
industrial sector shows considerable consumption of both natural gas and petroleum
products, and NEMS-RFF suggests that little opportunity exists for substitution between
natural gas and petroleum products in the industrial sector. Moreover, the baseline
NEMS-RFF projections show the U.S. industrial sector continuing to shrink in importance.
Over the 2008 to 2030 time horizon, the model projects that U.S. gross domestic product
(GDP) will grow at an annual rate of 2.45 percent per year from 2008 to 2030 while the
industrial sector grows at a slower rate of 1.37 percent per year.
In contrast with the lack of response that NEMS-RFF shows, Huntington (2007) finds
that natural gas consumption in the industrial sector is sensitive to differentials with
petroleum product and crude oil prices over the long run. An important part of the fuel
switching between natural gas and petroleum products occurs in the planning process when
firms are determining which fuel would be preferable to use. Similar issues may be
important in the commercial and residential sectors, where natural gas might be substituted
for heating oil if additional natural gas infrastructure were developed.
The combination of projected slow growth in the industrial sector and the relatively
weak substitution between natural gas and petroleum products in the industrial sector likely
contributes to the weak growth of natural gas consumption found in the baseline case. From
a policy perspective, the most important consequence of projections for slow growth in
natural gas use is likely to be the potential effects on LNG imports.
From the point of view of understanding how more abundant natural gas supplies
might affect the U.S. energy mix and policies to reduce CO2 emissions, the lack of
sensitivity in the industrial, residential, and commercial sectors of the NEMS-RFF model to
differentials between
19
Resources for the Future Brown, Gabriel, and Egging
natural gas prices and those for petroleum products may be an important issue. 24 If the
potential substitution of natural gas for petroleum products is greater than is assumed in
NEMS-RFF, natural gas prices could be considerably higher than the model projects, and
the effects of enhanced natural gas availability on U.S. energy markets could be much more
profound than our analysis shows.
5.2 Potential for Bottlenecks in Natural Gas Pipelines Brown and Yücel (2008a) show
that capacity constraints in the U.S. interstate pipeline
system have affected natural gas price differentials between various regions of the country.
The NEMS-RFF model allows pipeline capacity constraints to emerge between
interregional natural gas markets. In the model, pipelines cannot collect a congestion fee,
but regional price differentials do reflect pipeline capacity constraints when they occur.
Without probing the model to determine the extent to which pipeline constraints play
a role in the projections, it is difficult to assess how well the model reflects these market
realities. To the extent that the model fails to capture the full extent of congestion, it could
lead to more optimistic projections of natural gas use than are warranted. This issue may
be important under any of the scenarios that show rapid growth in the use of natural gas,
as is shown in Scenario 5.
5.3 Alaska Natural Gas Production Depending on which scenario is examined, the
NEMS-RFF model shows a natural gas
pipeline from Alaska to the lower 48 beginning operations as early as 2021 or as late as
2025.25 The earlier date is associated with Scenario 3, which incorporates more
conservative estimates of shale gas resources and a policy to restrict carbon emissions. The
later date is associated with Scenario 2, which incorporates more optimistic estimates of
shale gas resources and no policy to restrict carbon emissions.
The outcomes in NEMS-RFF assume that the decision to build the pipeline
depend strictly on the economics of North American natural gas markets. In contrast,
the various
24 Greater flexibility in the model with respect to
technology and fuel choice when equipment is retired and replaced would yield much stronger responses to
price differentials and greater energy efficiency in response to higher energy prices in the industrial, residential,
and commercial sectors. (See Section 3.1 above.)
25 Constraints on the model to limit the earliest potential date for the Alaskan pipeline to 2020 do not appear
to be binding.
20
Resources for the Future Brown, Gabriel, and Egging
competing proposals to build pipelines from Alaska are subject to a political approval
process, which is barely underway for one proposed pipeline. Moreover, one of the
proposals would result in Alaska natural gas being exported to Asia as LNG. And if natural
gas remains priced at par with oil in Japan (as is the current practice), analysis with the
World Gas Model shows that the combination of world oil prices and North American
natural gas prices projected with NEMS-RFF makes the export of Alaska natural gas the
most attractive option.
Delays in pipeline development or the redirection of Alaska natural gas to the export
market would reduce natural gas availability in the U.S. lower 48 market in the 2020s. Such
a development would increase the price of natural gas in the lower 48 market and make
natural gas less available as a bridge fuel to a low-carbon future. Failure to take into account
the possibility of U.S. natural gas exports should be considered quite a limitation of the
NEMS-RFF model.
5.4 Adequacy of U.S. LNG Import Terminals Currently, the United States has eight LNG
import terminals with a total capacity of 4.2
trillion cubic feet per year. The Federal Energy Regulatory Commission lists another six
terminals in the United States that have been approved and are under construction that
would add another 2.9 trillion cubic feet in annual capacity and one in Canada that would
add 0.4 trillion cubic feet. Another 14 terminals have been approved. Relative to the
projected LNG imports, which reach a height of 1.42 trillion cubic feet in 2018 under
Scenario 1, more than adequate LNG import capacity seems to be available.
5.5 The Adequacy of NEMS-RFF Of the issues we examined for NEMS-RFF, the
projected growth of the industrial sector,
26the substitution between natural gas and petroleum products in the industrial sector, and
the potential for U.S. exports of LNG all may make the modeling framework less than ideal
for examining how more abundant natural gas might affect U.S. energy markets. Pipeline
congestion is a lesser issue, and the adequacy of LNG import facilities does not appear to
be an issue over the projected time horizon.
26 Some other models better capture natural gas
market issues but without examining the complete energy system in which natural gas is used. (See Appendix
B.)
21
Resources for the Future Brown, Gabriel, and Egging
6. Uncertainties about U.S. Natural Gas Markets In examining natural gas markets, a
number of issues not explicitly considered in our
analysis may arise. Such issues include uncertainty about shale gas resources, the
possible elimination of oil and gas tax preferences, potential changes in access to
moratoria lands in which natural gas production is currently limited, the potential for
increased production of coalbed methane, and the potential availability of gas hydrates.
All of these issues affect the uncertainty about the outlook for U.S. natural gas markets.
6.1 Shale Gas Uncertainty and the Transition to a Low-Carbon Future The extent of
shale gas resources and their likely supply profiles remain highly uncertain.
In addition, EPA is taking steps toward regulating the hydraulic fluids that have been
important to enhancing the production of shale gas and boosting the estimates of
recoverable resources from shale gas formations (Obey 2009). This uncertainty about
shale gas resources and their potential development has important implications for policy.
Some policies—such as those mandating the use of specific technologies—require
accurate predictions about future resource availability and require technology change to be
costeffective. Policies that provide carbon pricing—such as cap-and-trade systems or
carbon taxes— do not. With pricing, market participants have an incentive to seek out the
most cost-effective means for reducing CO2 emissions, which makes such policies robust
across different projected scenarios. As the result of the considerable uncertainty about
shale gas supplies, the adoption of specific policies that favor or disfavor natural gas could
prove to be out of touch with market realities.
6.2 Elimination of Oil and Gas Tax Preferences The Obama administration has
proposed the elimination of what it identifies as oil and
gas company tax preferences (Office of Management and Budget 2009; U.S. Department of
Treasury 2009). The elimination of such preferences is consistent with an agreement
reached at the September 2009 G-20 summit in Pittsburgh for all G-20 countries to
eliminate fossil fuel subsidies (G-20 2009). Our analysis does not address this policy
change, but Allaire and Brown (2009) explain that these tax preferences account for less
than 1 percent of oil and gas revenue and show that eliminating the tax preferences would
have relatively small effects on U.S. natural gas markets.
According to Allaire and Brown, the elimination of the tax preferences would boost
consumer prices for natural gas 2.0 cents per million Btu in 2011 and 4.1 cents in 2030.
At the
22
Resources for the Future Brown, Gabriel, and Egging
same time, producer prices would see reductions of 3.2 cents and 4.3 cents below
baseline in 2011 and 2030, respectively. These price changes are relatively small when
compared to the projected prices for 2011 and 2030, which are $5.48 per million Btu in
2011 and $8.81 in 2030, respectively.
Allaire and Brown also estimate correspondingly small changes in market quantities.
They estimate that U.S. natural gas consumption will fall by 3 billion cubic feet below
baseline in 2011 and 49 billion cubic feet per year in 2030. They also estimate that domestic
natural gas production will fall 11 billion cubic feet below baseline in 2011 and 51 billion
cubic feet in 2030. Natural gas imports will rise 7 billion cubic feet above baseline in 2011
and 2 billion cubic feet in 2030. These quantities are relatively small in comparison to the
projected consumption figures for 2011 and 2030, which are 22.1 trillion cubic feet and 24.1
trillion cubic feet, respectively.
6.3 Moratoria Lands For the United States, current estimates of recoverable natural gas
resources are 1,760–
2,100 trillion cubic feet. Of this amount, about 162 trillion cubic feet is beneath federal lands
on which drilling has been restricted or off limits. These restricted areas are found in
Alaska, the Rocky Mountains, the Gulf Coast, and Appalachia.
In addition, another 92 trillion cubic feet of offshore natural gas resources are
unavailable for development, including 86 trillion cubic feet in the federal outer continental
shelf (OCS) moratoria regions. The OCS numbers are subject to considerable uncertainty
because estimates for some of the areas were made 25–40 years ago (NPC 2007). The
estimates could be increased with new exploration and assessments taking into account
modern drilling and extraction techniques.
In general, one would expect that increased access to these areas formerly excluded
from exploration and development would boost U.S. natural gas supplies. Such an effect is
likely to be stronger in the scenarios with higher natural gas prices because exploration and
production costs are generally higher in the moratoria lands. Consequently, one might
expect that increased access to moratoria lands would reduce some of the uncertainty about
U.S. natural gas supplies.
6.4 Coalbed Methane Coalbed methane is another unconventional natural gas source, and
engineering advances
in dewatering coal seams have boosted production since the 1990s (NPC 2007). As a
result, coalbed methane contributed 2.0 trillion cubic feet to U.S. natural gas production in
2008 (EIA
23
Resources for the Future Brown, Gabriel, and Egging
2009). Scenario 1 projects annual production of coalbed methane varying from 1.6 to 2.0
trillion cubic feet from 2009 to 2030, and NEMS-RFF shows production of this
unconventional resource to be price sensitive. Obviously, improved technology for coalbed
methane would increase the availability of natural gas in U.S. markets. For now, it is
probably safe to ignore the possibility of such technology changes.
6.5 Gas Hydrates Gas hydrates represent another unconventional natural gas resource,
but unlike coalbed
methane, gas hydrates are not yet in production. Gas hydrates are crystalline solids
consisting of gas molecules, usually methane, each surrounded by a cage of water
molecules. They have physical properties similar to ice and are found under permafrost in
Arctic regions and near the seafloor on continental slopes and in deep seas and lakes.
Resource assessments suggest that gas hydrates may be more plentiful than all other
carbon fuels combined, but future production technology and costs are extremely uncertain.
A substantial reduction in the costs of producing gas hydrates could increase the availability
of natural gas in markets worldwide. For now it is probably safe to ignore the possibility of
such technology changes.
6.6 Canadian Exports The Canadian and U.S. natural gas markets are intertwined, with
Canada a net exporter of
natural gas to the United States. The links between these two countries affects the
uncertainty in the outlook for U.S. natural gas markets. The potential for Canada to boost its
natural gas production reduces the uncertainty of the U.S. supply.
Canada has a large potential production area, the Mackenzie Delta, close to Alaska.
Its remoteness has been a hindrance in its development, but the depletion of other
production fields closer to the consumption and export markets has led to exploratory
activities. In recent projections, the Canadian National Energy Board (2009) estimates that
production will start in the Mackenzie Delta in 2017 at 0.29 trillion cubic feet and reach 0.44
trillion cubic feet in later years.
The Canadian National Energy Board projections extend through 2020 and show
Canada exporting 2.5 trillion cubic feet of natural gas to U.S. markets, including re-exports of
0.5 trillion cubic feet of LNG. In contrast, NEMS-RFF puts net imports of natural gas from
Canada at only 1.2 trillion cubic feet in 2020. The differences between these two projections
suggest that Canada could provide backup supplies of natural gas should U.S. natural gas
production fail to meet
24
Resources for the Future Brown, Gabriel, and Egging
expectations. It also suggests the possibility of a North American natural gas market
awash in supplies.
6.7 Mexican Imports In contrast with Canada, the Mexican natural gas market is only
weakly linked to the U.S.
market, with a small amount of pipeline exports from the United States to Mexico and Baja
California used for LNG imports destined for California. Mexico is also a potential
competitor with the United States for imported LNG. The projected growth of Mexican
natural gas consumption (and how it is met) increases uncertainty about the availability of
natural gas resources to the United States.
Until the early 1990s, Mexico was self-sufficient in natural gas. Since 2000, however,
Mexico has imported significant quantities of natural gas to meet rapidly rising domestic
consumption. Imports have ranged from 15 to 20 percent of consumption since 2004 (BP
2009).
Looking forward, the Mexican Ministry of Energy (Secretaría de Energía 2008)
projects that Mexican natural gas consumption will continue growing and domestic
production will plateau in 2010 with imports continuing to rise. NEMS-RFF shows slightly
lower Mexican imports, but the two projections move together. Nonetheless, the
composition of the expected imports differs according to the two sources, with the Mexican
Ministry of Energy projecting that all of its imports will be met by LNG by 2015, and
NEMS-RFF projecting increasing flows from the United States to Mexico via pipeline.
Over the shorter term, an outcome closer to the Mexican outlook for LNG imports
would increase the availability of natural gas in the United States, although that may mean
that Mexico is paying higher prices for LNG than it would for natural gas imported from the
United States. Over the longer term, Mexico presents the potential for rapidly rising natural
gas demand, which could reduce availability and boost prices in U.S. markets.
7. Summary and Conclusions To examine how natural gas supplies affect the
implementation of policies to reduce CO2 emissions, we compare five scenarios that
address different perspectives on natural gas availability, the availability of other resources,
and climate policy. We implemented these scenarios with NEMS-RFF and ran them through
2030. A comparison of the scenarios shows how the relative abundance or scarcity of
natural gas supplies might affect the use of natural gas as a bridge fuel to a low-carbon
future.
25
Resources for the Future Brown, Gabriel, and Egging
We find that abundant natural gas supplies increase use in most sectors of the
economy but do nothing by themselves to create a bridge to a low-carbon future. Without a
carbon policy in place, abundant and inexpensive natural gas fosters greater energy
consumption and displaces the use of nuclear and renewable resources to generate electric
power. Even though coal and oil use fall, the result is higher CO2 emissions. In contrast, if a
market-based policy, such as a cap-and-trade system, is in place to reduce
CO2 emissions, abundant supplies of natural gas can make a contribution in the transition
to a low-carbon future. With abundant natural gas resources, the estimated price of CO 2
emissions in a cap-and-trade system is slightly lower and policy implementation is less
costly. Moreover, the share of electric power generation from natural gas is greater in 2030
than in 2010 and much greater than with less abundant natural gas resources.
Other factors and policies could affect the use of natural gas. If the use of nuclear
and renewable energy for electric power generation develops more slowly than is expected,
abundant natural gas could prove more important as a bridge fuel in low-carbon policies. On
the other hand, the use of intervening mandates—such as renewable portfolio standards—in
conjunction with a cap-and-trade system can substantially alter market outcomes, reduce
the use of natural gas, and increase the costs of reducing CO2 emissions. From a broader
perspective, our analysis finds that the most cost-effective means for
reducing CO2 emissions depend greatly on projected resource availability and technology
changes—both of which are uncertain. If policymakers are to develop cost-effective policies
for controlling CO2 emissions, they must accurately predict the future or develop policies
that are robust across different projected futures. Economic theory suggests that pricing
schemes—such as cap-and-trade systems or carbon taxes—are robust in finding the most
cost-effective means for reducing CO2 emissions, regardless of how technology evolves or
how much natural gas or other energy resources are available.
26
Resources for the Future Brown, Gabriel, and Egging
Tables and Figures
Location Daily capacity Table 1. U.S. LNG Import Terminals (May 2009)
billion cubic feetAnnual cap acity trillion cubic feet Source: Federal Energy Regulatory Commission.
Everett, MA 1.035 0 .378 Cove Poi nt, MD 1 .800 0
.657 Elba Island, GA 1.200 0 .438 Lake Charles, LA 2 .100 0
.767 Gulf of Mexico 0 .500 0 .183 Offshore Boston 0.800 0
.292 Freeport, TX 1.500 0 .548 Sabine, LA 2 .600 0 .949 Total
4 .210
27
Resources for the Future Brown, Gabriel, and Egging
Table 6. Natural Gas Use in 2030, by Selected
Sector
S
c
n
r
o
1
e
a
i
trillion cubic fe
et Scenario 4
trillion cubic fe etTable
5. Electric Power Generation in 2030, by Sele
r
i
,305 Coal 2 ,311 1 ,306 2 ,223 1 ,070 Nuclea
Source: NEMS-RFF projections. trillion cubic feet Diffe
rence Total 23.46 25.32 7 .9% Electric power
.47 26.1% Industrial natural gas use 6 .29 6.47
.8% Commercial natural gas use 3.43 3
.39 –1.1% Residential natural gas use 4.87 4 .7
Table 7. Natural Gas Use in 2030, by Selected Secto
c
n
r
o
4
S
e
i
NEMS-RFF projections. trillion cubic feet Diffe rence
a
Source:
Total 25.32 26.53 4
.8% Electric power generation 8.47 9 .79 15.7% Industrial natural gas use 6
.47 6.36 – 1.8% Commercial natural gas use 3.39 3
.36 –0.8% Residential natural gas use 4.77 4 .73 –0.7%
28
Resources for the Future Brown, Gabriel, and Egging
29
Resources for the Future Brown, Gabriel, and Egging
Figure 4. Shale Gas Resources, Lower 48 States
30
Resources for the Future Brown, Gabriel, and Egging
Figure 6. Tra nsCan ad a’s P r op osed Al aska Pipeli ne
31
Resources for the Future Brown, Gabriel, and Egging
Figure 7. Mackenzie Delta
32
Resources for the Future Brown, Gabriel, and Egging
33
Resources for the Future Brown, Gabriel, and Egging
34
Resources for the Future Brown, Gabriel, and Egging
35
Resources for the Future Brown, Gabriel, and Egging
36
Resources for the Future Brown, Gabriel, and Egging
37
Resources for the Future Brown, Gabriel, and Egging
Appendix A. The Five Scenarios Scenario 1 (BAU–Low Gas) represents business as
usual with estimates of U.S. shale
gas resources at 269.3 trillion cubic feet. This case is based on EIA’s Annual Energy
Outlook 2009 as revised in April to include energy provisions in the stimulus package, but it
pulls new corporate average fuel economy standards forward from 2020 to 2016. (This
case is also known as NEMS-RFF Core 1.)
Scenario 2 (BAU–High Gas) represents business as usual with higher estimates of
U.S. shale resources. It is based on Scenario 1 with PGC estimates of U.S. shale gas
resources at 615.9 trillion cubic feet.
Scenario 3 (CO Policy–Low Gas) represents the implementation a low-carbon policy with
the low estimates of U.S. shale gas resources. It is based on Scenario 1 with a
cap-and-trade policy with CO22 emissions targets similar to those in the Waxman–Markey
bill and to those proposed by the Obama administration prior to the U.N. climate conference
in Copenhagen. (This case is also known as NEMS-RFF Core 2.)
Scenario 4 (CO2 Policy–High Gas) represents the implementation of a low-carbon
policy with the high estimates of U.S. shale gas resources. It is based on Scenario 3
with the higher estimates of U.S. shale gas resources.
Scenario 5 (CO2 Policy–High Gas–Limited Alternatives) represents the implementation of
a low-carbon policy with high estimates of U.S. shale gas resources and limits on the use of
nuclear power, carbon capture and sequestration (CCS), and renewable power generation.
These limits restrict the use of nuclear power to no more than 130 gigawatts by 2030, the
use of CCS to 2 gigawatts, and renewable power generation to no more than is found under
Scenario 4. Scenario 5 is based on Scenario 4 with the restrictions described.
38
Resources for the Future Brown, Gabriel, and Egging
Appendix B. A Comparison of Natural Gas Market Models Six models of natural gas markets play
a prominent role in the academic literature and/or
are used by industry. These models include the Rice World Gas Trade Model (RWGTM), World Gas
Model (WGM), GASTALE, ICF’s Gas Market Model (ICF GMM), GASMOD, and GRIDNET. In what
follows, we briefly examine the main characteristics of these models, including represented market
players, geographical coverage, and regional and periodical density. These comparisons reveal that
NEMS-RFF provides the most detailed representation for North American natural gas markets, the most
interaction with other fuels, and/or the appropriate time scale for analysis.
Capacity
expansions
Sectors
Seasons
Density
Time Scale
Number of
nodes
Market
power
Region(s)
Type
Table B-1. Summary of Natural Gas Model Characteristics
2030 Yearly 2 5
b Endogenous
Model
WGM MCP World Yes 41 2030 5years 2 3 EndogenousRWGTM CGE World No
NEMS-RFF LP USA+Canada No 15a
5years 1 1 Endogenous
c MCP Europe+LNG Yes 6 2025 10years 1 1 EndogenousGASTALE MCP
GASMOD
39
2030 5years 3 3 EndogenousGRIDNET LP USA No 18,000 Operational M
Exogenous
ICF GMM LP? USA No 114 Several years Monthly 12 4 Exogenous
Notes: CGE, Computable General Equilibrium; LP, Linear Program; MCP, Mixed
Problem.
a United States 12, Canada 2, and Mexico
1.
Includes electric power generation, which is not considered an end-use sector in
NEMS.
b
c
The dynamic version of GASMOD.
RWGTM and WGM provide much less detail than NEMS-RF
American market (Table B-1). The principal advantage of these mo
coverage, which allows the models to better capture the interaction
markets in different world regions. WGM also addresses the poten
development of international market power, but it does not allow fo
supply and demand conditions in a detailed bottomup approach tha
changing economic conditions.
GASMOD and GASTALE also address market power aspects expl
coverage is strictly European when it comes to demand. Gridnet an
U.S.
Resources for the Future Brown, Gabriel, and Egging
coverage but are designed to support short- to medium-term business decisions. Neither
is well suited for long-term scenario analysis.
In comparison with some other models, the two principal issues that NEMS-RFF does
not address well are the interaction of the North American natural gas market with the rest of
the world and the market power aspects that may arise in global natural gas supply.
Because current projections show the North American market as nearly self sufficient, these
issues are probably not too important. To the extent that any scenarios examined show
sizable imports of LNG, these issues may deserve more attention.
Rice World Gas Trade Model According to Hartley et al. (2004a and 2004b), Hartley and
Medlock (2005), and Brotzen
(2007), the Rice World Gas Trade Model (RWGTM) is a model of international natural gas
markets developed at Rice University with Market Builder software from Altos Management
Partners. RWGTM is a dynamic spatial equilibrium model. Periods are linked through the
optimal scheduling of production over time. The demand side is modeled via a bottom-up
approach, using econometric estimates based on GDP and population size as well as prices
of gas and competitive fuels. Agents at the supply side of the model maximize their
discounted profits, based on expected prices and demand levels.
The model includes the exploration of new reserves and the construction and
expansion of transportation options and determines market-clearing prices to balance
demands and supply. In a perfectly competitive setting, producers maximize their
discounted profits. The model develops and expands pipeline and LNG transportation
infrastructure.
RWGTM takes a hub-and-spoke approach to LNG transportation. Market power
aspects are not explicitly accounted for. The time horizon of the model is 2050, and the
model data set includes more than280 demand regions and 180 supply regions.
World Gas Model According to Gabriel and Egging (2009) and Egging et al. (2009), the
World Gas Model
(WGM) is a multiperiod, mixed-complementarity model for the global natural gas market,
allowing for endogenous capacity expansions in the LNG, pipeline, and storage sectors.
The model contains 41 countries and regions and covers about 98 percent of worldwide
gas consumption and production, both with seasonal variation. The model includes a
detailed representation of border-crossing natural gas pipelines and contractual and spot
trades in LNG.
40
Resources for the Future Brown, Gabriel, and Egging
The represented market agents include producers and traders as competitive
players; liquefaction, regasification, pipelines, and storage as regulated players; and three
final consumption sectors—power generation, industry, and residential/commercial. Traders,
in the pipeline as well as in the LNG market, can exert market power relative to the end-user
markets. The model base year is 2005 and covers the period up to 2040 in five-year steps.
Future production capacities and consumption curves are scenario-dependent input
parameters, whereas capacity expansions in transportation and storage are endogenously
determined by model agents. The model is equipped to represent a global cartel in the world
natural gas market.
GASMOD According to Holz et al. (2008) and Holz (2009) GASMOD is a two-stage gaming
model
for the European upstream and downstream gas markets developed by DIW Berlin. The first
stage is a noncooperative game wherein producers determine their production and export
levels to European countries. In the second stage, traders determine their deliveries to
consumer markets within Europe. Pipeline and LNG import capacities limit the transportation
of gas to and within the continent. The geographical coverage of the model is Europe and
exporters to Europe. Two model versions are available. The static model allows for market
power on two levels (double marginalization): producers/exporters to traders and traders to
end users. It contains 13 exporting countries and 17 importing regions. The dynamic model
contains the three largest exporters to Europe and the three largest importers in Europe.
The dynamic model assumes a perfectly competitive upstream market and includes
endogenous capacity expansions. The distinguished market players are producers, traders,
consumers, pipelines, and LNG regasification. The period up to 2025 is represented in three
10-year periods. No seasons or sectors are distinguished.
GASTALE According to Lise and Hobbs (2009), GASTALE is a gas market model for the
European
gas market developed by the Energy Research Center of the Netherlands. GASTALE is a
mixedcomplementarity model with endogenous investments by transmission and storage
system operators. The model solves for a single-year short-run equilibrium in each five-year
period for a number of consecutive periods. In between each period solution, a separate
routine decides on capacity expansion levels for pipelines, liquefaction, regasification, and
storage, based on the expected congestion prices (generated with an open-loop information
structure). Investments should bring the expected capacity congestion prices below an
acceptable threshold. The
41
Resources for the Future Brown, Gabriel, and Egging
represented market agents include producers, consumers, and transmission and storage
operators. Producers can exercise market power relative to end-user markets. Transmission
and storage are assumed to be regulated players. Production capacity expansions and
development of the demand curves are exogenous to the model. The model has three
seasons and three final consumption sectors: power generation, industry, and
residential/commercial.
The regions represented in the model include Europe, the gas-exporting neighboring
regions, and some representative LNG exporters. Of the 19 regions, 7 serve only as
consumers, 9 only as producers, and 3 as both producers and consumers. The model is
solved in five-year intervals through 2030.
GRIDNET According to Brooks and Neill (2010), Gridnet models a large network for the
United
States. It contains more than 18,000 nodes and 200,000 pipelines. The model is very similar
to a huge, generalized, minimum-cost flow problem with losses and some complicating
constraints. It is a decision support tool that simultaneously addresses supply, demand, and
the transportation of natural gas for gas trading companies. The detail in the representation
of the pipeline network supports the operational planning problem of companies, such as the
routing of existing contract volumes, as well as analyzing opportunities and setting prices for
new contracts.
The model has two modes: an operational mode and a planning mode. Many
contractual, capacity, and other constraints add operational restrictions; therefore, the size
of the operational model is an order of magnitude smaller than the whole model, and it can
be solved in little time. This is further facilitated by the pre-computation of costs and losses
for origin–destination node pairs. As long as the pipeline system does not have bottlenecks,
the pre-computed values remain valid. When used for planning, several of the constraints
are relieved, in particular the joint constraints for minimum and maximum aggregate
deliveries at sets of destinations. This allows for a more global optimization approach at the
cost of much higher calculation times, using special solvers rather than general LP.
Market power is not addressed in Gridnet. The (implicitly) distinguished market
players are producers, consumers, and pipelines. The model is not a market equilibrium
model, and sectors and seasons are not represented as such. However, the time
resolution of the model and the supply contracts that companies have to meet (existing) or
those they are investigating (opportunities) imply a high level of detail.
42
Resources for the Future Brown, Gabriel, and Egging
ICF’s GMM According to ICF International (2009 and 2010), ICF GMM determines monthly
marketclearing prices for supply and demand in the United States and Canada. The model
investigates the impact of different assumptions, such as those regarding supply sources
and demand drivers on prices, produced volumes, and transportation flows.
The model contains 114 nodes, of which 77 are supply nodes. Most U.S. states are
represented as single nodes; however, larger states are represented as several nodes.
Other nodes represent the Canadian provinces, offshore pipeline landing points, border
crossings with Canada and Mexico, the Alaskan LNG export terminal, and all LNG import
terminals. The represented market players are producers, consumers, pipelines, and LNG
terminals. Market power is not addressed.
Supply curves, which represent production from existing and new fields, imports from
Mexico, and LNG regasification, address costs for storage and pipeline use. An optional
add-on is a hydrocarbon supply model that provides more detail for gas production. GMM
includes four demand sectors: power, industry, residential, and commercial. A separate
electricity market module projects natural gas consumption for power generation. The
demand curves for other sectors are driven by weather (degree days) and economic output.
LNG exports and exports to Mexico are also represented on the demand side of the model.
Transportation capacities are exogenous to the model.
Appendix B References Brooks, Robert E., and C.P. Neill. 2010. GRIDNET:
Natural Gas Operations Optimizing
System.http://rbac.com/Articles/GRIDNETNaturalGasOperationsOptimizingSystem/t
abi d/67/Default.aspx (Accessed March 2010).
Brotzen, Franz. 2007. Economists’ World Gas Trade Model Gives a 30- to 50-year
Estimate of Natural Gas Supplies, Demand, Prices. Rice University, Houston, Texas
http://www.media.rice.edu/media/NewsBot.asp?MODE=VIEW&ID=9253&SnID=2
(Accessed March 2010).
Egging, Ruud, Franziska Holz, and Steven A. Gabriel. 2009. The World Gas Model: A
MultiPeriod Mixed Complementarity Model for the Global Natural Gas Market. Discussion
Papers of DIW Berlin 959, DIW Berlin, German Institute for Economic Research
Gabriel, S.A., and R. Egging. 2009. World Gas Model. Unpublished paper. University
of Maryland.
43
Resources for the Future Brown, Gabriel, and Egging
Hartley, Peter, Kenneth B. Medlock, III, and Jill Nesbitt. 2004a. Rice University World Gas
Trade Model. James A. Baker, III Institute of Public Policy, Rice University, Houston Texas
(March).http://www.rice.edu/energy/publications/docs/GSP_WorldGasTradeModel_Part
1_05_26_04.pdf (Accessed March 2010).
Hartley, Peter, Kenneth B. Medlock, III, and Jill Nesbitt. 2004b. Rice University World
Gas Trade Model. James A. Baker, III Institute of Public Policy, Rice University, Houston
Texas (December).
http://www.forum.rice.edu/presentations/Forum04/Peter%20Hartley%20%20Presentation
%20%20An%20Economic%20Model%20of%20the%20Gas%20Industry.pdf (Accessed
March 2010).
Hartley, Peter, and Kenneth B. Medlock, III. 2005. The Baker Institute World Gas Trade
Model. James A. Baker, III Institute of Public Policy, Rice University, Houston Texas.
http://www.rice.edu/energy/publications/docs/GAS_BIWGTM_March2005.pdf (Accessed
March 2010).
Holz, Franziska, Christian von Hirschhausen, and Claudia Kemfert. 2008. A Strategic
Model of European Gas Supply (GASMOD). Energy Economics 30: 766–788.
Holz, Franziska. 2009. Modeling the European Natural Gas Market—Static and
Dynamic Perspectives of an Oligopolistic Market. Ph.D. Dissertation, Technische
Universität, Berlin, Germany.
ICF International.2010. Gas Market Model
http://www.icfi.com/markets/energy/doc_files/nangasweb.pdf (Accessed March 2010).
ICF International. 2009. GMM, Model Overview. Lise, W., and B.F. Hobbs. 2009. A
Dynamic Simulation of Market Power in the Liberalised
European Natural Gas Market. The Energy Journal 30(special issue
1):119–136.
44
Resources for the Future Brown, Gabriel, and Egging
References Allaire, Maura, and Stephen Brown. 2009. Eliminating Subsidies for Fossil
Fuel Production:
Implications for U.S. Oil and Natural Gas Markets. Issue Brief 09-10.Washington,
DC: Resources for the Future.
Böhringer, Christoph, and Knut Einar Rosendahl. 2009. Green Serves the Dirtiest: On
the Interaction between Black and Green Quotas. Discussion paper 581.Kongsvinger,
Norway: Statistics Norway.
BP. 2009. Statistical Review of World Energy. London, UK: BP. Brown, Stephen P.A.,
and Mine K. Yücel. 2008a. Deliverability and Regional Pricing in U.S.
Natural Gas Markets. Energy Economics 30(5): 2441–2453. Brown, Stephen P.A.,
and Mine K. Yücel. 2008b. What Drives Natural Gas Prices? The Energy
Journal 29(2): 45–60. EIA (Energy Information Administration). 2009. Annual
Energy Outlook 2009 (revised).
Washington, DC: U.S. Department of Energy. EIA (Energy Information
Administration). 2010. Annual Energy Outlook 2010 (preliminary).
Washington, DC: U.S. Department of Energy. G-20. 2009. Group Statement on
Pittsburgh G-20 Summit. Leaders’ Statement. The Pittsburgh
Summit. September 24–25, 2009.
http://thepage.time.com/group-statement-onpittsburgh-g-20-summit/
(accessed November 19, 2009).
Hartley, Peter R., Kenneth B. Medlock III, and Jennifer E. Rosthal. 2008. The
Relationship of Natural Gas to Oil Prices. The Energy Journal 29(3): 47–65.
Huntington, Hillard G. 2007. Industrial Natural Gas Consumption in the United States:
An Empirical Model for Evaluating Future Trends. Energy Economics 29: 743–759.
Krupnick, Alan J., Ian W.H. Parry, Margaret A. Walls, Kristin Hayes, and Anthony C.
Knowles. 2010. Toward a New National Energy Policy: Assessing the Options. Washington,
DC: Resources for the Future.
National Energy Board. 2009. 2009 Reference Case Scenario: Canadian Energy Demand
and Supply to 2020—An Energy Market Assessment. Calgary, Canada: National Energy
Board.
45
Resources for the Future Brown, Gabriel, and Egging
Navigant Consulting. 2008. North America Natural Gas Supply Assessment. (July.)
Chicago, Illinois: Navigant Consulting.
NPC (National Petroleum Council). 2003. Balancing Natural Gas Policy.
(September.) Washington, DC: NPC. http://www.npc.org/ (accessed March 12,
2010).
———. 2007. Facing the Hard Truths about Energy. (July.) Washington, DC:
NPC. http://www.npchardtruthsreport.org/ (accessed March 12, 2010).
Obey, Doug. 2009. Rulings Strengthen NEPA Oversight of Gas Drilling Absent SDWA
Rule. Inside EPA Weekly Report 30(37). September 18.
Office of Management and Budget. 2009. A New Era of Responsibility: Renewing
America’s Promise. Washington, DC: Executive Office of the President of the United
States.
Palmer, Karen, and Richard Sweeney. 2010. Modeling Policies to Promote Renewable and
Low Carbon Sources of Electricity. Washington, DC: Resources for the Future.
PGC (Potential Gas Committee). 2009. Potential Supply of Natural Gas in the United
States: Advance Summary. (June.) Golden Colorado: PGC.
Secretaría de Energía (Mexico). 2008. Natural Gas Market Outlook 2008–2017.Mexico,
D.F, Mexico: Secretaría de Energía.
U.S. Department of Treasury. 2009. General Explanations of the Administration’s Fiscal
Year 2010 Revenue Proposals. Washington, DC: U.S. Department of Treasury.
Villar, Jose, and Fred Joutz. 2006. The Relationship between Crude Oil and Natural Gas
Prices. EIA Manuscript (October). Washington, DC: U.S. Department of Energy.
Weyant, John P. 2009. Overview of Energy Modeling Forum Study 22: Climate Change
Control Scenarios. Energy Modeling Forum. Stanford, CA: Stanford University.
46